Connected Architecture.
Structured, secure communication protocols link edge intelligence, industrial control, and monitoring — each doing exactly one job, with no single point of failure.
Secure control-layer communication
The AI decision engine and the industrial controller communicate over a secure industrial protocol, authenticated with certificate-based identity verification and end-to-end encryption. This channel carries sensor telemetry, control signals, and the heartbeat signal used for fail-safe monitoring — protected against tampering and replay.
Lightweight telemetry distribution
Real-time telemetry — including the live Risk Index and processing performance metrics — is published over a lightweight publish-subscribe messaging protocol. This decouples the AI processing layer from the monitoring layer, so multiple dashboard clients can subscribe to live data without adding load back onto the AI unit.
Protocol summary
| Protocol | Endpoints | Purpose |
|---|---|---|
| Secure industrial protocol | AI unit ↔ Controller | Encrypted control data, telemetry, heartbeat |
| Lightweight messaging | AI unit → Monitoring node | Real-time telemetry publishing |
| TCP/IP (Ethernet) | All devices | Underlying local network layer |
| Email delivery | Monitoring node → Email | Critical alert notifications |
| Cellular network | Monitoring node → SMS / Call | Alerts independent of local network availability |
Why it is separated this way
Splitting communication into distinct, purpose-built channels means a slowdown or failure in one layer — say, the monitoring dashboard — cannot propagate into the control layer responsible for physical safety actions.